云环境下面向复杂资源需求的虚拟机能效部署研究

针对云环境下在线虚拟机部署这一矢量装箱问题进行了研究,提出了多维空间划分模型和在线虚拟机能效部署算法OEEVMP。多维空间划分模型可以引导虚拟机部署,避免多维资源的不均衡利用;基于此模型,提出的OEEVMP算法在物理机运行数量局部最优和全局最优之间取得均衡,从而提高虚拟机部署能效。通过仿真实验,将OEEVMP算法与MFFD算法进行了对比,实验结果验证了所提算法的可行性和有效性。最后,对控制模型的两个参数进行了分析,给出了最佳的参数组合。

Research on Energy Efficient Virtual Machine Placement with Complicated Resources Requirements in Cloud

Online virtual machine placement problem in cloud which is also a bin packing problem is studied in this paper. Multi-dimensional space partition model and online energy efficient virtual machine placement algorithm are presented. The proposed model can be used to guide the placement of virtual machines and avoid the imbalanced use of multi-dimensional resources. Based on this model, the proposed OEEVMP can achieve a tradeoff between local optimum and global optimum with the number of running physical machines, so as to improve the energy efficiency of virtual machine placement. The proposed algorithm is also evaluated with MFFD algorithm via extensive simulations and experiments. The experiment results verifie the feasibility and validity of the proposed algorithm. At last, two parameters of the model are analyzed and the optimal combination of them is given.